If white LEDs are actually blue, then why hasn't anyone tried green?

[shao.fu.tzer]

From my understanding, a white LED is just a blue LED with phosphor to change the emitted light white. Well, every spec sheet I look at shows that colored green LEDs are more efficient and brighter than their equivalent blue partners. Has anyone tried adding phosphor to green LEDs? Or maybe a doped phosphor material that would convert the green wavelength to white. It would greatly enhance efficiency... I don't know if this has been discussed, but it seems like someone should be doing some research on this. Is it because the difference in spectrum from green to white so different compared to blue to white that it becomes inefficient?
Just curious...

Shao

 

[Kinnza]

Because green LEDs have way lower efficiencies than blue ones. There is a lot of research trying to improve them, but currently internal quantum efficiencies of green LEDs are below 20%, compared to 90% of blues.

So it is way more efficient to convert blue light with phosphors to green than using green LEDs.

 

[shao.fu.tzer]

Because green LEDs have way lower efficiencies than blue ones. There is a lot of research trying to improve them, but currently internal quantum efficiencies of green LEDs are below 20%, compared to 90% of blues.

So it is way more efficient to convert blue light with phosphors to green than using green LEDs.

Rebel Blue - 58lm @ 700mA
Rebel Green - 166lm @ 700mA
Cree XR-C blue - 18.1lm @ 350mA
Cree XR-C green - 51.7lm @ 350mA
Cree XP-E blue - 39.8 lm @ 350mA
Cree XP-E green - 107lm @ 350mA
Cree XR-E blue - 30.6lm @ 350mA
Cree XR-E green - 67.2lm @ 350mA

I could go on... I don't see how blue LEDs are more efficient... please elaborate on quantum efficiencies...

 

[Harold_B]

The LEDs based on deep blue (405 - 470nM) don't actually convert the blue to white. The phosphor typically has a peak output centered around 560 - 575nM for a cool white with a skewed distribution that has less output in the red than the green end of the spectrum. If you do a search for YAG you can find several phosphors with published spectrums for both the emission and absorption and there should be a spec for peak excitation. What the eye perceives as white is actually just the yellow-green output and unabsorbed blue combined.

To make a white LED from a green source would require a phosphor that has a peak excitation in the green spectrum and has an emission spectrum in the red or orange. You would also still need to add the deep blue to complete the spectrum to make a reasonable white or else what you would get is a sickly green. The other issue of having a phosphor that has an emission spectrum that is close to the excitation wavelength is that you will be reabsorbing the already down converted light which will be very inefficient.

The other option might be to use an organic phosphor (a dye) but there are processing and heat issues that make lifetime short for organics. I know there are labs working on those types of phosphors and they are already available from big companies like BASF and Merck plus smaller laser dye makers. The patent minefield is treacherous in the organics just like the inorganics.

I hope that makes some sense since I'm trying to keep a reply on a really complex topic short. That and I know just enough to be dangerous....

 

[Harold_B]

You are perceiving the lumen conversion as efficiency when that is not the case. The human eye is most sensitive to green and the lumen is a measure of the eye's perception of brightness. If you were to measure two devices of different out wavelengths with the same source die, the correct way to compare them would be in mW. That takes the eye's sensitivities out of the calculation. This a link that goes into greater depth explaining how you can calculate the reason why the green LEDs have a seemingly higher efficiency:

http://www.optics.arizona.edu/palmer/rpfaq/rpfaq.htm#radiometry

 

[shao.fu.tzer]

Well thanks for that explanation... It DOES makes sense... But luminous efficacy is luminous efficacy... Right?

 

[slebans]

Well thanks for that explanation... It DOES makes sense... But luminous efficacy is luminous efficacy... Right?

The Luminous Efficacy of Radiation(LER) value of monochromatic light at 555nm is 683 lumens per watt. If the LED was 100% efficient and you powered it with 1 watt then it would emit 683 lumens.

The Luminous Efficacy of Radiation(LER) value of monochromatic light at 455nm is 33 lumens per watt. If the LED was 100% efficient and you powered it with 1 watt then it would emit 33 lumens.

You might want to read this thread:
http://www.candlepowerforums.com/vb/showthread.php?311740-LEDs-waste-75-as-heat


Stephen Lebans

 

[IMSabbel]

Well thanks for that explanation... It DOES makes sense... But luminous efficacy is luminous efficacy... Right?

Well, what slebans said, plus: like ***

take a look at the curve here:
http://en.wikipedia.org/wiki/Luminosity_function

To get a luman at 550nm, you need only a few % of the photons required at 420nm.

But you say that the lm/W of the greens are already so very good? Well, if you would use them to create white light, efficiency would _drop_, as any other wavelength has less lumen/W. Plus you would lose energy in the conversion.

And last, but not least, you still need blue light, and it is very difficult to create blue light from green ones (in fact, totally impractical for LEDs)

 

[csshih]

Well thanks for that explanation... It DOES makes sense... But luminous efficacy is luminous efficacy... Right?

well yes, but to put it in really simple terms what we now need to look at is now much emitted radiation (like Harold said, in mW) we get from the emitters.
let's hop to lasers for a sec(as they're rated in mW) - a 1W 532 (your standard green) laser would be visually much brighter than a 1W 445 (blue-violet, common now 🙂 ) laser. They will, though, have identical burning strength (disregarding absorption spectrums/focus and the such)

Craig

 

[bshanahan14rulz]

One of the biggest constraints, and probably the reason we don't just slap some phosphor on top of vcsels, is heat. I believe remote phosphor tech. reduces this problem, and that is probably how the BMW headlights everyone's talking about work. Instead of heat having to be wicked away through an active thermal heat source, through the LED's substrate, through the MCPCB, and to the heatsink, the phosphor could be directly mounted on a heatsink and overdriven a bit more, possibly creating even higher surface brightness.

IIRC, GE tried using Violet emitters under their phosphors, I believe they were also dabbling with remote phosphor domes, but not for better heat dissipation.

 

[csshih]

One of the biggest constraints, and probably the reason we don't just slap some phosphor on top of vcsels, is heat. I believe remote phosphor tech. reduces this problem, and that is probably how the BMW headlights everyone's talking about work. Instead of heat having to be wicked away through an active thermal heat source, through the LED's substrate, through the MCPCB, and to the heatsink, the phosphor could be directly mounted on a heatsink and overdriven a bit more, possibly creating even higher surface brightness.

IIRC, GE tried using Violet emitters under their phosphors, I believe they were also dabbling with remote phosphor domes, but not for better heat dissipation.

Hehe, yep, everyone was pretty much going "omg, laser beam headlights" sheesh, marketing.

anyways, I believe the original Lux Vs were actually UV emitters with phosphor.

Craig

 

[MikeAusC]

Well thanks for that explanation... It DOES makes sense... But luminous efficacy is luminous efficacy... Right?

But when you're getting light through ILLUMINATING PHOSPHOR, Luminous Efficacy becomes a useless number.

 

[shao.fu.tzer]

Well thanks for the explanations everyone... I've been reading thoroughly into the subject...
So I understand about the lost efficiency through phosphor - but in my original post I did mention
possibly finding another substance besides YAG phosphor that could be used? I wasn't talking about
actual present day technology - more of the possibility of squeezing white light from a green LED.

Hypothetically speaking, couldn't a different, obviously undiscovered scintillator, efficiently convert the green
wavelength to white?

From what I'm getting from this - lumens ratings as we know them are more about perceived light to an individual, rather than actual measured output? If this is the case, why do people even bother with lumens ratings? 300 lumens to me, could only be 200 lumens to you if you were less sensitive to the same spectrum of light? Is that right?

Sorry if I'm a little slow on this one, I've always focused on the electronic/electrical and mechanical aspects of lights and am definitely not a scientist...

 

[csshih]

Well thanks for the explanations everyone... I've been reading thoroughly into the subject...
So I understand about the lost efficiency through phosphor - but in my original post I did mention
possibly finding another substance besides YAG phosphor that could be used? I wasn't talking about
actual present day technology - more of the possibility of squeezing white light from a green LED. From what I'm getting
from this - lumens ratings as we know them are more about perceived light to an individual, rather than actual measured
output? If this is the case, why do people even bother with lumens ratings? 300 lumens to me, could only
be 200 lumens to you if you were less sensitive to the same spectrum of light? Is that right? Sorry
if I'm a little slow on this one.

I'm sure humans will all vary (but there really is no way to tell, as we all at least see consistently.)
in this case lumen output is useful, as all we care about it usable *visually* detectable light.
(wheras in many cases of lasers all you want is energy output). Like HaroldB said, Radiometry versus Photometry.

Craig

 

[shao.fu.tzer]

Hmmmm... looks like I should have used Google before starting this thread... I just did a Google search and discovered that apparently someone has done just what I was talking about... although you need a paid subscription to read the whole document...

Phosphor-Free White Light From InGaN Blue and Green Light-Emitting Diode Chips Covered With Semiconductor-Conversion AlGaInP Epilayer


Edit: I just read an interesting PDF on the nrel.gov site that covers the same topic. Do search a for:
"Green Light-Emitting Diode Makes Highly Efficient White Light" - there are also a few references to
a phosphor-free white LED that's in development.

 

[csshih]

Edit: I just read an interesting PDF on the nrel.gov site that covers the same topic. Do search a for:
"Green Light-Emitting Diode Makes Highly Efficient White Light" - there are also a few references to
a phosphor-free white LED that's in development.

Had a look at that doc. it's just RGB color mixing?

Craig

 

[MikeAusC]

. . . From what I'm getting from this - lumens ratings as we know them are more about perceived light to an individual, rather than actual measured output? If this is the case, why do people even bother with lumens ratings? 300 lumens to me, could only be 200 lumens to you if you were less sensitive to the same spectrum of light? Is that right? . . . .

You're right - without an objective standard, it would be like trying to compare brightness of torches !

There is an Internationally accepted standard to convert radiant power to luminous flux - at 555nm green, there are 683 lumens per watt of radiated power.

 

[shao.fu.tzer]

Had a look at that doc. it's just RGB color mixing?

Craig

Looks as if, but they still say that the efficiency would be higher - and it would allow on-the-fly tint control which would be absolutely awesome, if you ask me. No more warm vs. neutral vs. cool - just twist a knob until you get a color that you're happy with. I see great things in the future! 😀

You're right - without an objective standard, it would be like trying to compare brightness of torches !

There is an Internationally accepted standard to convert radiant power to luminous flux - at 555nm green, there are 683 lumens per watt of radiated power.

That sounds like an awful lot... I say - go green! 😀 But really though... I hope all of this is leading somewhere. I got lost about 5 posts back! 😀

Shao