[UVC SAFETY WARNING] First tests and experiments with a 265 nm UVC LED

[Stefano]

Should he worry for the 185 nm line? For what concerns eyes and skin, he will be protected wearing usual protection.

 

The main problem I guess it's ozone. The 185 nm line is so energetic it splits O₂ molecules which can then recombine to form O₃. From your measurements it seems this line is quite weak, but I can't really say anything since I don't have experience with these lamps.

 

David, do you smell ozone when using it?

[dabateman]

I hadn't noticed much ozone in the past when using it. keep the filter on and you should be ok. My filter fully covers the bulb. But sadly doesn't seem to block much. The 405nm and 435nm are visible. As are the 550nm and 580nm lines.

 

So I have know idea what that is. It might be 1mm zwb3. Or some plastic doped to try and block some visible but failing.

[enricosavazzi]

I think these LEDs make it practical to begin exploring the realm of UVC-induced fluorescence in the UVA. To detect this fluorescence, just use these LEDs to illuminate the subject and the usual UVA imaging kit with Baader U filter etc. This will filter out any fluorescence in the VIS and NIR.

 

EDIT: The 4-chip variant will allow a much more compact light source. The copper "star" PCB should be able to handle 4W. https://www.ebay.com/itm/265nm-Deep-UVC-3535-LED-Diode-6V-12V-24V-4Chips-for-led-UV/233642453371?hash=item36662ce97b:g:2w4AAOSwalxfBCDj

[bvf]

I think these LEDs make it practical to begin exploring the realm of UVC-induced fluorescence in the UVA. To detect this fluorescence, just use these LEDs to illuminate the subject and the usual UVA imaging kit with Baader U filter etc. This will filter out any fluorescence in the VIS and NIR.

 

That will be one of the things I will be trying when I get my 254nm lamp, as well as looking for differences between 254nm and 365nm excitation for visible and IR fluorescence.

[Stefano]

Ok, to clarify and to help our readers too here’s the link for purchase: https://www.ebay.com/itm/254318067890 I will put it in the original post too.

 

I too was thinking about putting four chips on a star PCB, but I don’t know if they sell empty PCBs with that kind of pads. Buying one of the LEDs Enrico suggested to then replace the chips could work. With four chips you will have a power consumption of about 4 W and a power output (according to the data provided by the seller) of 32-40 mW. If you can put 9 chips you will have about 9 W in and 72-90 mW out.

 

UVC LEDs have some big disadvantages (very low efficiency and very low power compared to common mercury lamps) but also all the advantages LED technology offers. Fully dimmable to your taste, instant on/off, relatively clean and tight spectrum, small size and easy to operate. I agree with Enrico on that if you are a beginner these LEDs could help you start experimenting with UVC light in a relatively safe way. These LEDs are very small, so if you don’t want them around anymore you can just put them away in a little drawer, while mercury lamps are significantly bigger.

[dabateman]

Well the cost has finally dropped below $50, where it was less than 2 years ago for a single 265nm led chip.

It looks like 4 leds on a board are common. I see multiple Ebay ads for that. Tricky bit is I am not sure what driver they use. Enrico's ad has one listed, but I don't know if that will work with the 4 chip units. I am also confused why there are multiple input voltage options. With double the voltage leading to half the amperage, but will their driver work with that?

 

Its almost at the cost worth experimenting, as I have wanted to do UVC induced UVB and UVA. Problem with Mercury bulbs is there advantage and you get all the UV lines. So discrete detection isn't possible. But thats great for reflectance detection as the filter is on the camera.

A germacidal bulb seem ok at offering most ranges with my filters, except the 302nm line and weak 395nm line. But others are strong.

 

UV is really in the hands of most photographers these days.

 

 

[Stefano]

One 265 nm LED has a forward voltage of about 6 V. The energy of a 265 nm photon is ~4.68 eV, so, theoretically, that should be the forward voltage of those LEDs, and also the pretty wide bandgap of the semiconductor used. (BTW, to be sure to give the appropriate credits, Andy and I "discovered" this correlation independently).

 

Having said that, I think they offer three configurations: 4P (all chips in parallel), 2S2P and 4S (all chips in series). This will result in forward voltages of 6, 12 and 24 V. Having all chips in series ensures the current flowing through each LED is the same, and you have more control, but 24 V is quite a high voltage for small LED drivers, but not that difficult to obtain.

 

I too wanted to find a suitable driver to build my 265 nm UVC flashlight, I searched everywhere for a driver with 3.7 V input (Li-ion battery) and fixed 150 mA output current and capable of 6-7 V, but I found nothing. The best I found is a 50 W buck converter with adjustable current, but its power rating makes it much bigger in size than what I need and, being a buck converter, I also need to boost the voltage from the battery. Not easy.

 

About $10 for one 1 W 265 nm LED is not bad. $50 for five. Still a bit expensive, but affordable for a small project.

Thorlabs LEDs are surely high quality, no doubt on that, and also more powerful, but really too expensive. The ones I have seem to work well, and the glass opacity and the banknote fluorescence indicate they truly emit UVC.

 

About UVC-induced UVB/UVA fluorescence (UVCIUVBF/UVCIUVAF, not an elegant abbreviation), I think it is definitely possible. I observed something similar with my 340 nm LED: https://www.ultravioletphotography.com/content/index.php/topic/3770-340-nm-led-first-impressions

 

Jonathan once said this is used on skin for medical purposes: https://www.ultravioletphotography.com/content/index.php/topic/3753-fire-dive-gear-underwater-fluorescence-photography-equipment/page__view__findpost__p__34015

 

One additional problem you may encounter with UVC LEDs, related to this, is that the chips themselves may emit some UVB and UVA from UVC self-induced UVB/A fluorescence. Typical UVA LEDs (like the now common 365 nm chips) emit a whitish/yellowish light (bluish with Nichia chips) for the same reason. I have a chinese 10 W 365 nm LED, the 9 chips type, which has a strong vanilla-colored emission when turned on. If I shine another (filtered) 365 nm UV flashlight on it, I observe the same glow.

 

The same may happen to UVC LEDs. I observed a dim glow in the visible spectrum with my protected eyes, from violet to red, it probably keeps going in the NIR region and I am pretty sure it also goes down to the upper UV region. Filtering away this light isn't easy. 330-type glass passes all UV from UVC to UVA to some violet. You need a dichroic filter if you want a very clean spectrum, but I have to say it is already quite clean, similar to an unfiltered Convoy torch.

[Andy Perrin]

I think these LEDs make it practical to begin exploring the realm of UVC-induced fluorescence in the UVA. To detect this fluorescence, just use these LEDs to illuminate the subject and the usual UVA imaging kit with Baader U filter etc. This will filter out any fluorescence in the VIS and NIR.

 

EDIT: The 4-chip variant will allow a much more compact light source. The copper "star" PCB should be able to handle 4W. https://www.ebay.com/itm/265nm-Deep-UVC-3535-LED-Diode-6V-12V-24V-4Chips-for-led-UV/233642453371?hash=item36662ce97b:g:2w4AAOSwalxfBCDj

At least there will be no arguments about white balance!

[dabateman]

 

At least there will be no arguments about white balance!

 

True 313nm is green, 360nm is yellow and 390nm is blue. No discussion or arguments.

[Stefano]

Better not to start discussing that as I already know it would be problematic. I personally would just use a normal UV white balance and maybe a custom white balance to my taste to remove color casts if present.

[Cadmium]

I think Pylon used one of these Triple Wavelength lights for his UV(A/B/C)IVF shots.

https://www.polmanmi...y_too_cool.html

Just as a comparison. His UVA induced shots may have been done with torches or flash, not sure.

 

By the way, here is Pylon's UVIVF page.

https://photoextremist.com/ultraviolet-induced-visible-fluorescence-photography-tutorial

[Stefano]

While doing the green banana test with these LEDs, I burned one out because I shorted the power supply with the copper PCB and the resulting voltage spikes (there is probably an inductor in series with the output inside the power supply) killed it. It works, but it is partially shorted, and I saw one of the bond wires glow orange. Fortunately I have other four of them.

 

When I took a new one, I noticed it was a bit different. Not the LED itself, but the visible color it produced. The old one was bluish, this one appears more pink/magenta. So there is some variability, but the LED behaved the same.

 

Also, I think I found other LEDs which look like identical to the ones I bought, in case the old listing is not working anymore: https://www.ebay.com/itm/293632351696