I use 99.9% Isopropyl Alcohol on my SFT40 and it doesn’t dissolve the phosphor.
I did however push too hard once and i broke part of the Phosphor off.
Not always the case; I’ve cleaned different types of LED’s with Iso, after a reflow or soldering wires to make sure there is no flux on the LED, and most of the time I had no problems.
As Nachtfeuerzeug and INeedMoreLumens stated: I’ve also cleaned an SFT-40 (only once so far) with Iso and had no problem afterwards.
Not sure though what happened on DesertSnake his example, besides it seems he had (very) bad luck…
3 Thanks
Is the phosphor on the SFT40 held in place by some sort of binding/adhesive agent, or just cohesion between the grains like a dried chunk of clay? If the latter, it might explain the weakening of the structure when wet.
Really wish they considered some sort of protective measure during manufacturing, like a thin silicone layer with the Cree high-intensity emitters, a glass layer like the 519A, or whatever Osram does.
I’ve cleaned SFT-40 with IPA and I wondered did I just imagine or did the tint change. Then I tried again and it definitely changed to unusable.
After learning that lesson, I forgot I learned it, and did the same thing to another specimen ruining it.
For my defence, there was wide time gap between those mistakes.
3 Thanks
Pure blue, it is no different from the blue LED on the SF Helmetlight. I would be more precise, but I don’t have a spectrometer. My camera sees a little purple, I only see blue
QReciprocity42
Pursuit of brightness, additional layers = additional losses, I guess
3 Thanks
Well, beer sure will add a little tint and if you drank some, too, there could even be more perceptual differences. 
Sadly not that kind of IPA
1 Thank
Where’d you buy it from?
I guarantee there’s knockoffs going around.
1 Thank
I don’t think it’s fake. I always bought from Simon and everything was perfect
https://aliexpress.ru/item/1005006178451351.html?sku_id=12000036697197978
Hmm, what CCT was it?
I ordered 5000K, my friend tested it in M6LT, everything corresponds
Good to know. Ive cleaned a bunch of 6500k sft40s with iso and never had a problem, but never a 5000K. That’s interesting
1 Thank
In this regard it could be interesting that my (one and only) 5000 K sample was totally burned while testing the maximum luminous flux. The phosphor layer was burned, similar to the problem with the XHP50.3 HI earlier. The 6500 K and also the 3000 K ones are totally fine, even at maximum current for several minutes.
Maybe there is something really different between the variants.
2 Thanks
I fixed it a little, found 9 pcs 219b 5000K, 1 sst20 4000k, 1 xml2 3000k, cut off the phosphor from them, put them on top and secured them with transparent glue. These 219b 2011 were low CRI and had a terrible beam with a hole in the center
It’s like if I were a medieval ruler, took a bunch of virgin girls, drained their blood into a bathtub and climbed in there to continue living. Or rather, someone cut all my skin off, took a bunch of innocents, cut their skin off and put it on me as new skin. This is just like Frank Cotton from Hellraiser 1
9 Thanks
This is amazing!! I really wish this LED can be sent and tested lol, unless you already have some equipment to estimate its performance.
Now imagine transplanting high-CRI R9080 phosphor on an extreme thrower like the SFT40, or a very efficient flooder like XHP50.2.
2 Thanks
I already thought about this and also imagined some sort of semi automated desk-sized small production line for very own custom made LEDs 
(with 3D printing and less expensive CNC this is not as unrealistic as it seems)
But the main problem is to get the raw phosphor mixture and the proper silicone/epoxy coating, which is needed here. Salvaging 12 LEDs for just one emitter is way too expensive.
5 Thanks
I looked at the prices for phosphors and was horrified, thousands of dollars for few gramms 
And so you came up with a great idea, if the blue peak is approximately the same from different manufacturers, then this method can work, but need thinner equipment and a cleaner place to make the modification efficiently. I would send you this diode, no problem at all, but now parcels have become very expensive, I usually send something abroad in a large pile to one of friends and then he sends it to other places
1 Thank
This is great. You found yourself in a hole and kept digging, and somehow ended up on top of the mountain
2 Thanks
This would be a sick project!
I like this idea because many high CRI LEDs (like the Moonleds previously tested) are limited in performance by the underlying blue die. But as you both pointed out, sourcing the phosphor in a cost-effective way is a limiting factor. Please keep the diode–I do not have the appropriate equipment to perform measurements, but thank you for offering!
I have recently been messing with UV resin as an optical adhesive, specifically using them to re-dome dedomed or sliced LEDs. So far the resin seems to be stable under the flux density and temperature of an operating LED (no yellowing after a month), the main weakness is that the resin can detach during reflow, due to thermal stress. I think LEDSupply used to sell “LED Seal” which is a silicone sealant designed for this type of stuff; if one were to perform these modifications on a large scale, the adhesive won’t be the limiting factor, but the phosphor.
I’ve accidentally de-phosphored 2 519As, and detached the entire die on another one, while trying to pry off the resin. But the nice thing is that the entire phosphor comes off intact with the glass layer, and I could try to transplant the phosphor on a high-performance die to see what happens. I’ve got a 50.2 but it would take 4 519A dies to cover; I do have a random collection of 3535 LEDs but they are fairly outdated.
I tried to build a PC green LED by using green glow powder as the phosphor, but the powder is quickly overloaded by the flux density, going dim then angry blue. The resin seemed fine.
1 Thank
I am curious, what glue are you using? I’m surprised to see it holding up fine, at least for the moment. Maybe I’ll get some to experiment with.
Also what is the cat-shaped object in the first photo?
1 Thank