What would be the best choice for shooting in UVC 200-300nm range?

[gavin1911]

Hi folks, I am new to ultraviolet photography and I appreciate your inputs and thoughts. I am searching for a digital camera, filters, and a LED light source that together are good for UVC 200-300nm. The camera could be a converted DSLR or anything else that has more than 1 Mega-pixel.

 

:)

[Andy Perrin]

You are asking the impossible (or very nearly). Nobody here shoots UVC. It is also SUPER DANGEROUS. Even getting to UVB is difficult with our equipment. Nearly every single obstacle is stacked against you: the lens will not transmit unless it is quartz ($$$), the camera cannot be a consumer camera because they have no sensitivity to UVC (and very little to UVB without debayering in some cases). Most UVC filters have very low transmission and you need to block out of band light very well or you would have contamination.

 

I believe there are some companies that make UVC cameras but you can expect to pay tens of thousands of dollars or the equivalent.

[dabateman]

Basically what Andy said.

There is a work around depending on what you want to image. You can make or buy a phosphorus screen that converts the lower uvc to higher Uva or visible for detection.

Thats what the cheaper $5000 cameras do.

 

If your into Diy, you could build a scan back type camera using a spectrometer sensor. You can get the sensors cheapish off ebay and follow the design that people have done with scanners to large format cameras. But you will need to use nothing but fused silica.

 

There are 260nm and 280nm Leds you can find off ebay or Alibaba as well. But very expensive. You can use the 254nm very low mercury bulbs to get something. But UVC isn't really about imaging more about detection.

 

As most things will be black down there. Protein absorb at 280nm, DNA at 260nm and peptide backbones absorb at 230nm. So don't expect to see much but black.

 

It does come back to what do you want to photograph?

 

Only things I can think of you would see down there would be molten metal, explosives and a transformer explosion. That limits you to gun ranges and blacksmiths. Or a blacksmith with a gun?

[Andy Perrin]

Don’t forget power line ionization and candle flames! But seriously I suspect the original poster simply didn’t realize that UVC is probably not what he really wants to image. He said something about being interested in using it for forensics which would not be smart for either the photographer or the subject unless they want nasty burns.

[Dmitry]

I know only devices like UV-260 corona camera for inspecting long HV power lines.

[Cadmium]

Arch welding.

 

https://aoemj.biomedcentral.com/articles/10.1186/s40557-015-0087-7

Scared yet?

[dabateman]

Candle flame, perfect.I forgot about that. I for some reason was fixed on hydrogen burning.

I could safely image that using my full spectrum Em1 and my 260bp10 filter to let you know how you will not see anything. I don't think the Em1 goes that low. But I do think I got to 280nm imaging the sun directly. The exposure of the sun was 10 seconds and clouds appeared in those images that I couldn't see.

 

I am waiting for Andrea's response about don't do this you will kill yourself as UVc is extremely dangerous. And she is right! Don't even think about using a 254nm light source without taking all the proper precautions. I could also try imaging my 254nm light, but I don't like being around. I will have to configure the wifi Olympus app to remotely control the camera and be out of the room. But an image of a bulb seems boring. A bouncing candle flame sounds more fun.

[gavin1911]

Thank you all for your comments. I am now becoming more scared and concerned about safety issue - I found a few vendors with 1-2mW LED bulbs in 250-280nm range ( For example, this one manufactured by SETi) - would these LED bulbs also be dangerous when being pointed to human?

[Alaun]

Well, the data sheet your link points at says:

 

CAUTION, UV LIGHT: These devices emit high intensity ultraviolet (UV) light, which is harmful to skin, eyes, and may cause cancer. Avoid exposure to UV light, use protective glasses and cover your skin.

[Andy Perrin]

All UVC is extremely dangerous to humans, not just in the long term but in the “You will instantly hurt yourself” sense.

[JMC]

Gavin, the main issue is how much light you'll need, given the low sensitivity of the cameras that far down. I'd be very surprised if those leds are giving much light out, so you'd need loads and I presume they aren't cheap. If you can talk about it, please give more information on what you hope to achieve and perhaps we can suggest ways to go about it.

[gavin1911]

Well, the data sheet your link points at says:

CAUTION, UV LIGHT: These devices emit high intensity ultraviolet (UV) light, which is harmful to skin, eyes, and may cause cancer. Avoid exposure to UV light, use protective glasses and cover your skin.

 

 

this is damn scary now...perhaps I should try 300-400nm?

[gavin1911]

Gavin, the main issue is how much light you'll need, given the low sensitivity of the cameras that far down. I'd be very surprised if those leds are giving much light out, so you'd need loads and I presume they aren't cheap. If you can talk about it, please give more information on what you hope to achieve and perhaps we can suggest ways to go about it.

 

For the project, it's used to detect Fentanyl which has a distinctive absorption profile in the 200-300nm range from Heroin and Cocaine. This will be in close range (<5inch) and small area (<1 sq.inch and a thin layer).

[Cadmium]

Everyone else has already said everything.

300nm (320nm) to 400nm is all you can really expect.

List:

1) full spectrum converted camera - limited to 300nm ~ 280nm? max if pushed to extremes, big maybe, monochrome full spectrum would help but... not enough.

2) most lenses will not transmit below 320nm max (many much less), special expensive lenses will transmit lower, so the limit there is money.

3) sub 300nm filters are available, but the point is already mute.

4) UV-C light is available and quite dangerous. Arch welders do it every day, but every inch of their body is covered by thick protective gear impervious to the UV.

 

"CAUTION, UV LIGHT: These devices emit high intensity ultraviolet (UV) light, which is harmful to skin, eyes, and may cause cancer. Avoid exposure to UV light, use protective glasses and cover your skin."

The WARNING you quoted is pertinent to ALL UV (UV-A, B, and C). Light used for even UV-A (320nm - 400nm) has the same warning, however, it becomes extremely serious with UV-C/B.

Never be around any UV-C light source, like sterilizing bulbs, etc., including arch welding.

We commonly use 365nm LED light, full spectrum flash, and other light that contains "high intensity UV" mostly in the UV-A range, and even it is dangerous.

You don't just walk around with a UV-C bulb and shine it at stuff, quite dangerous and unpopular.

 

You are ultimately limited by the camera sensor.

Unless someone here knows of some fancy expensive sub-300nm camera, then I don't know how you can do it.

[JMC]

For the project, it's used to detect Fentanyl which has a distinctive absorption profile in the 200-300nm range from Heroin and Cocaine. This will be in close range (<5inch) and small area (<1 sq.inch and a thin layer).

Ok, so what will be in the background? Do you expect to be able to image against a variety of backgrounds, or will you be collecting a sample and then imaging it against the same background evert time. That far down, most organic materials will be absorbing most of the UV, and the world is pretty dark, so how are you expecting to differentiate the absorption of what you want to see from everything else?

 

As you've gathered by now there are some huge hurdles to overcome if you need to do straight reflectance imaging that far down - not only equipment but safety. Would fluorescence imaging be an option?

[dabateman]

Your experiment reminds me of circular dichroism. It has the power to give you full protein structure. However, the signals are broad and the background is very high, (many things will give a similar signal). That its reduced to only giving insight into secondary structures, beta sheets or alpha helices.

 

You will have similar issue on location for imaging this molecule. You really will not know. Best to image in isolation, and then most likely best using a uv/vis spectrometer.

[gavin1911]

Ok, so what will be in the background? Do you expect to be able to image against a variety of backgrounds, or will you be collecting a sample and then imaging it against the same background evert time. That far down, most organic materials will be absorbing most of the UV, and the world is pretty dark, so how are you expecting to differentiate the absorption of what you want to see from everything else?

 

As you've gathered by now there are some huge hurdles to overcome if you need to do straight reflectance imaging that far down - not only equipment but safety. Would fluorescence imaging be an option?

 

Fentanyl has 0-20% absorption at certain wavelengths between 245-320nm - my plan is to take straight reflection images at these wavelengths and write an algorithm to process the images to differentiate fentanyl from cocaine/heroin.

[ulf]

Fentanyl has 0-20% absorption at certain wavelengths between 245-320nm - my plan is to take straight reflection images at these wavelengths and write an algorithm to process the images to differentiate fentanyl from cocaine/heroin.

 

Fentanyl has 0-20% absorption at certain wavelengths between 245-320nm - my plan is to take straight reflection images at these wavelengths and write an algorithm to process the images to differentiate fentanyl from cocaine/heroin.

 

If you consider using an UV-LED in that range, please be aware of that the spectrum of a LED normally is rather narrow and the optically emitted power is low.

 

I have no hands on experience with UV-LEDs below 365nm for UV-illumination.

From datasheets I can see that the optical power output is really tiny compared with LEDs at 365nm and above, as the efficiency is low.

Even for LEDs at 340nm the conversion efficiency is down to 3% compared to at around 30% for LEDs at 365nm

https://www.aptechno...s.co.uk/LEDs/UV

 

Most LEDs have a 50% peak width more narrow than 15nm and the available wavelengths are not that many.

I made some measurements of a series of 1W LEDs in and close to VIS that show the general behaviour.

[dabateman]

 

 

Fentanyl has 0-20% absorption at certain wavelengths between 245-320nm - my plan is to take straight reflection images at these wavelengths and write an algorithm to process the images to differentiate fentanyl from cocaine/heroin.

 

That sounds slightly more doable. Like a spectral unmixing experiment. Do you have access to a uv/vis spectrometer?

I would hit each with specific wavelengths, and look at their absorbing spectrum to see the tails. Unmixing these is how 7 color flow cytometry works.

 

[gavin1911]

That sounds slightly more doable. Like a spectral unmixing experiment. Do you have access to a uv/vis spectrometer?

I would hit each with specific wavelengths, and look at their absorbing spectrum to see the tails. Unmixing these is how 7 color flow cytometry works.

 

nah..I went to local police squad yesterday and they only have an HPLC in the lab..do you have any recommendation on the model or type of uv/vis spectrometer that I can use to generate an absorption spectrum for powders?

[JMC]

Gavin, sorry to be asking annoying questions (by all means tell me to go and have a beer), but if this is for law enforcement work, do you need to consider the specificity of any measurement - the degree of certainty that you are identifying what you think you are? Having worked in a lab environment for year, some things required GLP (good lab practice) some things require GMP (good manufacturing practice). GMP was for when we were making cosmetics for human testing, and only certain equipment could be used for making and testing of those products. GMP required a higher level of calibration and tracability than GLP. For law enforcement work, I'm assuming any analysis on the UV-Vis of the powders would need to be done to a certain level of GMP (or your equivalent) and as such that will have an impact on what equipment you can use.

[gavin1911]

Gavin, sorry to be asking annoying questions (by all means tell me to go and have a beer), but if this is for law enforcement work, do you need to consider the specificity of any measurement - the degree of certainty that you are identifying what you think you are? Having worked in a lab environment for year, some things required GLP (good lab practice) some things require GMP (good manufacturing practice). GMP was for when we were making cosmetics for human testing, and only certain equipment could be used for making and testing of those products. GMP required a higher level of calibration and tracability than GLP. For law enforcement work, I'm assuming any analysis on the UV-Vis of the powders would need to be done to a certain level of GMP (or your equivalent) and as such that will have an impact on what equipment you can use.

 

It's a good question. This project is intended to build a point-of-care screening device for drug checking in the field so as long as there's no false negative it's all good. False positive's okay cuz they can bring it back to lab to confirm. I've built algorithms for some other material/substances in NIR/VIS range in the past. My current concern's whether I can get an image, even very very low quality, under 200-300nm UV and whether I can find right equipment/components to put together a prototype.

[dabateman]

I have never worked with powders, so wouldn't know. Everything, I have taken into solution. I am not sure a powder would work for transmission type experiment in standard spectrometer. You would need a reflective one, if one exists. Something similar to a metallurgical microscope.

 

Now we are getting complicated again.

 

Ok they do exist:

https://www.hunterlab.com/powder-color-measurement.html

 

https://www.ssi.shimadzu.com/products/uv-vis-spectrophotometers/diffuse-reflectance-measurement.html

 

And I have done this once. Seeing the figure jolted my brain. In an analytical lab I remember making a fine paste and spreading it out on finely polished KCl disc and on other salt disc, that I can't remember now. The KCl disc was invisible to IR and worked to hold the same.

 

I don't think I would be much help in the correct or even finding an affordable instrument. But quick online search came up even with X-rite having devices.

[ulf]

I don't think I would be much help in the correct or even finding an affordable instrument. But quick online search came up even with X-rite having devices.

 

That depends of how you define affordable.

Do X-rite really reach UV-C?

https://oceanoptics....-transmittance/

With a suitable lightsource, spectrometer and bi-fucated fibre the method in the link above might be usable.

 

A spectrometer approach will likely solve much of the problems that a image sensor based method have, but it will give a "one pixel image" result.

[Andrea B.]

SAFETY NOTICE:

I'm not intending to interfere or stop the good discussion here. However, I must pop in as the Editor-in-Chief and remind everyone that UVP does NOT support UV photography between 200-300nm. That should only be done with extreme protection in a laboratory setting with specialized gear. The typical photo gear converted for UV work in the 330-400 nm range won't work in the 200-300nm range anyway.

 

Thank you for listening!!

Carry on.......