Ha, this is great to see!

I looked at the graph and my very first thought was "that doesn't look like a 405nm LED, it looks like a 395nm LED". And what do you know? This is especially good because it drops off earlier than the 405, leaving more data un-overlapped in the visible range, for fluorescence data. @matej - take a look!

I would like to find a source for cheap but strong 395nm LEDs. Dave or @stoft, know any? Extra points if it's 5v and can run on USB power. :-)

Dave, can you post links to your data, so I can plot it against the 405nm LEDs to see how wide the base of their peaks are?

Thanks, Dave!

Very consistent results, too, BTW! Although i see there is a change in baseline noise between your two laser tests.

Oh i see here from your post to the list:

https://spectralworkbench.org/sets/3176 UV flashlight 395nm baselines feb23 using .09mm slit https://spectralworkbench.org/sets/3177 UV laser 405nm baselines feb23 using .09mm slit

Here's my 405nm LED test with isopropyl compared with your 395nm/isopropyl test:

Your baseline noise is higher, so it's hard to tell whether the shorter wavelength LED gets us less overlap with the visible light region we want to use in oil identification. Both seem to bottom out at around 430-450... It's hard to tell, but 395 might bottom out about 10nm earlier, which would make sense. Do you know what brand of LED they use? @stoft, thoughts on the baseline noise here and what we can say about the comparison?

Hey Jeff, glad you found this useful, I believe the baseline noise is caused by the extra cell length (the glass test tubes are 2.05cm.) I am waiting for some quartz cuvettes I ordered to come in, which are the standard 1cm cell length, so I'll have to re-calibrate all my baselines when they arrive.

I'll see if I can find out the brand info on the UV flashlight LED's and the UV laser pointer, also this is the source I use to by a lot of parts, I like it because they give the tech docs on most of the parts you buy.

http://www.allelectronics.com/

They have a good selection and it's cheap!!

Here is the Peak and FWHM data for both the 395nm and 405nm readings:

PEAK DATA UV laser 405nm 1) Isopropyl alcohol wavelength [nm] absorbance FWHM [nm] 251.08 15.2000 -- 262.17 15.2000 -- 276.58 15.2000 -- 294.31 15.2000 -- 322.02 15.2000 -- 345.3 15.2000 -- 370.79 15.2000 -- 401.82 91.7800 2.9176 2) Ethanol wavelength [nm] absorbance FWHM [nm] 253.3 17.8000 -- 269.93 17.8000 -- 287.66 17.8000 -- 401.82 155.0000 3.1708 3) distilled water wavelength [nm] absorbance FWHM [nm] 402.93 155.0000 2.7092

UVflashlight 395nm distilled water (4) cell length 2cm

wavelength [nm] absorbance FWHM [nm] 271.04 0.2056 -- 289.88 0.1973 -- 323.13 0.1975 20.974 353.06 0.1759 -- 388.52 1.0140 27.334

UV flashlight 395nm Isopropyl alcohol (5) cell length 2cm

wavelength [nm] absorbance FWHM [nm] 279.62 0.1558 -- 295.99 0.1550 -- 394.19 0.9675 28.729

UV flashlight 395nm Ethanol (6) cell length 2cm

wavelength [nm] absorbance FWHM [nm] 271.98 0.1710 36.528 392.01 1.0140 27.957

Here is where you can buy 395nm UV LED's for $1.40 a piece.

http://www.allelectronics.com/make-a-store/item/uled-2/ultraviolet-led-395-nm-30-degree/1.html

http://www.allelectronics.com/make-a-store/item/led-92/purple-5mm-led/1.html - here is the 405nm violet LED for $1.10 a piece(there is also a PDF spec sheet on this.)

Ah, excellent, excellent, but it's my understanding that the plastic LEDs like that aren't typically as bright as the surface mounted flat LEDs I'm using here: https://publiclab.org/notes/warren/01-05-2016/testing-high-brightness-405nm-leds-in-fluorescence-spectroscopy-of-oils

Sorry to ask again, but do you know how to identify LEDs as bright as these?

Hey Jeff, yeah that's an ultra-brite LED (high current/high-flux output.) it's just mounted on a heat sink assmbly.

I have a few of those, but they are RGB LEDs and white.

The 395nm flashlight that I use are 5mm Ultra-violet LED's. Emits blue 395nm UV light. Water-clear lens. 3.7 Vdc, 20 mA. 30 degree beam pattern. and there are 51 of them, pwr suppy is only 3.0vdc. I could emulate the same thing you have there with only one ultra-brite blue or violet LED running on the same pwr supply.

here is a picture of a 3W LED mounted on a heatsink and an optional focal lens, this LED's wavelength is 410nm-465nm.

11401.Jpg

13779.Jpg

Hmm, i looked on DigiKey and found some more options, some high brightness, and one listed as 365nm, which'd be great:

http://www.digikey.com/short/33z3p7

One, the VLMU3500-385-060TR-ND, pictured here, looks very similar to what I'm using, but 385nm, 100x minimum order, $4.40 each:

It'd have to be soldered to aluminum, i'd guess. If it's as narrow as the ones we're using, that might get us clear data down to 425nm, instead of only above ~440-450nm.

The 160-2184CT-ND, at $4.70-$8, is at 365nm, and 500MA/665MW, which is just a little less power than the one I'm using. It runs on 4v, but says "none available". But i'd love to test it!

The 365nm might get us clear data above 410nm -- possibly better than the laser!

Mouser has that one! http://www.mouser.com/ProductDetail/Lite-On/LTPL-C034UVH365/?qs=pPiH5vUCJeR2s22VxlBhsw==

I think I should order it. I can solder it to one of the heat sinks I have -- i burnt out one of my LEDs.

Ok, this is getting a little out there, but I'm going to buy and try a few other products; I see

$3.84: http://www.ebay.com/itm/201481504603

$3.99: http://www.ebay.com/itm/UV-Ultraviolet-USB-LED-Light-Power-KeyChain-Lamp-Portable-Torch-Scan-Money-/291564049819

These may not be the right wavelength or power, but at these prices, I wonder if the companies which make them would be willing to do a custom run with the LEDs we've selected. Gosh, they're cheap!

Hey Jeff, I thought I had some high brightness LED's, I can't find them right now but I did have a RED LED clear top 1.7v@20mA 3000mcd@ 12deg viewing angle, I got a small proto board, 9V battery and a couple of resistors to fire it up and set it up with my focusing lens I use for my laser and below are the results of that, which looks like its not too bad.