[UVC SAFETY WARNING] Easy mod to get a UVC sensitive sensor?

[Lou Jost]

NEVER look at a UV-C light.

NEVER let UV-C light hit your skin or eyes directly or by reflection.

UV-C light can cause:

• severe burns of the eyes and the skin, and
• DNA damage from broken chromosomes.

When working with UV-C illumination, you MUST:

• cover up completely, 
• wear head & eye protection, and
• have strong ventilation.

 

---

 

 

I was thinking about UVC photography, and wondered if the Bayer filter dyes would fluoresce and cause problems. Then I realized this would not be a problem but a solution. If a layer of material on the sensor would fluoresce in UVC, then the photocells would record the pattern of light. What if we could coat the sensor (preferably de-bayered) with a thin layer of phosphors? Maybe some kind of paint, applied in very thin layers (preferably a mix with white fluorescence, if the sensor still has its Bayer matrix). Any irregularities in layer thickness could be fixed using the astrophotography technique of applying "flats" .

 

If the layer would be relatively opaque to visible light, then the sensor would only record UV fluorescence, not visible light, so with care we could dispense with UV-pass filters!

 

This technique is of course a variation of a standard technique for visualizing things like x-rays.  I wonder if maybe we could figure out a simple variation for our application.

[colinbm]

Interesting, has been past my tiny brain too.
There are phosphor film/sheet, but I haven't been able to get them to send samples, or purchase any.
I really would like to have some.
https://phosphortech.com/

[Andy Perrin]

Oh that was proposed many years ago. It should work in principle- they do it for SWIR even- but the caveats include very nonlinear wavelength response and of course you need to put the phosphor right on the sensor, not the cover glass or even the bayer, or risk light spreading to adjacent pixels, and internal reflections. You can also photograph a phosphor screen, which has already been done by Dabateman using the Sirchie fingerprint device that Jonathan, Dabateman, and I got on eBay. That device uses a phosphor in conjunction with an photomultiplier.
 

TJNelson proposed applying phosphor to the chip in 2013 or so on his webpage (which is full of outdated information so reader beware)

Quote

Another possibility is to use a quartz lens to focus the UV light onto a phosphor screen and photograph the screen with a regular camera. According to James R. Janesick, a yellow highlighting marker creates an efficient phosphor coating that can allow detection of UV light down to 50 nm. Almost down to X-rays, baby!

 

The phosphor that Janesick uses is Lumogen yellow S 0790 (2,2′-dihydroxy 1,1′-naphthalazine), aka Pigment Yellow 101. Unfortunately, most modern highlighters use Basic Yellow 40 or Acid Yellow 23 instead. For this to work, the dye must be applied directly to the chip. This means the protective glass must be removed—a very difficult and risky procedure. The procedure must be done with a torch or micro-milling tools under a microscope and is a specialized skill.

There’s also the issue that we often want to switch filters to capture the same scene in other wavelengths. So having a specialized camera has that versatility drawback.

 

A debayered raspberry pi with phosphor applied would be a way to test this, at a price.

[Lou Jost]

I think painting a thin layer of a phosphor (perhaps the yellow highlighter liquid) on a normal sensor could potentially be easier and better than photographing a phosphor screen. Nonlinearity of response is no big deal. Making the sensor useless for anything else is definitely a drawback. But given how hard it is to image extreme UV, it seems worth playing with....

[JMC]

I can confirm that drawing on a bare sensor with a highlighter pen does indeed increase UV sensitivity. However I did it in such an uncontrolled way that the image quality was rubbish. It boosted sensitivity on both a monochrome part of the sensor and a part with the Bayer filter still present.

A few months back I looked into getting a sensor coated with a fluorescent or phosphorescent coating. It is doable and a few companies do the process, but where it all fell down is that they were used to dealing with raw sensors, not ones attached to a board. Also, they weren't comfortable removing the coverglass and wanted it sending to them without it on, which would have added more expense and complexity. They can do the coating on a monochrome sensor or one with the Bayer filter attached. There is a downside of a hit to resolution though. In the end for the cost it just wasn't worth it.

Funnily enough I have seen an odd effect when I tried to image UVC with a camera which had a coverglass made from Schott WG280 which blocked the light. I got a ghostly image, and I think that was from the WG280 fluorescing under the UVC and the camera then picking that up. It was like an image taken through a window with mist on it.

[lukaszgryglicki]

Yeah, I wasn't sure if I ever will be able to make an UV-C image with my 50R, but I ordered quartz cover glass "just in case" - right now my biggest issue is:

1) have no time.

2) If I use 3 25W low pressure mercury bulbs at the same time, very close to target and then I try to photograph the subjest (lp mercury bulbs are supposed to emit over 70% of their output as UV-C) I'm probably recording only thir blue visible glow. I have Edmund filter supposed to pass only 253.7nm but it seems to be angle dependent/dichroic - I can see world outside through it with my eyes and I guess camera also can do it. Another Edmund filter is for 222nm light, it is pitch black but, obviously, not for mercury 253.7nm line. And my KrCl lamp (unfortunatelly from China) is broken, I cannot return it and I wasted $250 for it - it glows orange and makes buzzing sound - it doesn't seem to output any 222nm light (it glows in purple just for 1s after start up and then turns orange & makes that buzz sound).

 

Yet if I have more time (Christmas?) I will try to do all possible combinations of my UV lights & filters and maybe maybe maybe will see something in UV-V (I really hope).

 

EDIT: so I'm beginning to wonder that even if UV-C is able to hit my mono/quartz sensor directly - can it be registered then? Maybe there is still some glass on the "pixel"/"detector" level that absorbs that UV-C photons and they are NOT registered anyway? I just don't know yet.

 

My ultimate target now is to get images at 253.7nm line.

 

 

[Lou Jost]

Quote

My ultimate target now is to get images at 253.7nm line.

I have a gigantic special 254nm lens that I tried out last night, using a germicidal lamp with a major line at 253.7nm. I got nice images with a full spectrum Sony A7R, but I suspect the image was really done with the tiny amount of energy this lamp emits in UV-A. My low-pass UV filter probably cuts out the UV-C. I wish I could tell. Anyway, I'll make a separate post about that.

[lukaszgryglicki]

I know NO lowpass filter that passes UV-C, at least none of those described on forum - all of them block UV below 280 and leak IR

Edmund optics makes BP filters, but, as I said my 253.7nm one is dichroic and probably only works at angle = 90 degrees - I can see through it outside. Another 222nm one seems to be blocking everything (I see no leaks and camera sees no leaks) - but it also blocks mercury UV-C line and my KrCL lamp is not working :(

 

[Lou Jost]

51 minutes ago, lukaszgryglicki said:

I know NO lowpass filter that passes UV-C, at least none of those described on forum - all of them block UV below 280 and leak IR

Edmund optics makes BP filters, but, as I said my 253.7nm one is dichroic and probably only works at angle = 90 degrees - I can see through it outside. Another 222nm one seems to be blocking everything (I see no leaks and camera sees no leaks) - but it also blocks mercury UV-C line and my KrCL lamp is not working :(

 

 That's what I figured....Mine doesn't bleed IR though. Anyway the lens seems to work very nicely at whatever UV wavelength it is transmitting. The black flower centers are present.

[Andy Perrin]

Black flower centers appear right next to visible (only 380-400nm is enough for that). It’s not much of a test. Most lenses will pass that range, even modern ones with many elements. 
 

The history of people trying to photograph UV-B and UV-C on this board is full of examples of people initially fooling themselves with leaks because they haven’t used enough blocking. The sensor gain falls nearly exponentially in UV, so it’s really hard to know how much to use. So far everyone who has done it successfully (dabateman and Jonathan so far) progressed in stages with UV-B first, so they knew how much blocking was needed for that, then they moved to UV-C. So a multi-stage approach seems to be the best way. 

[Lou Jost]

40 minutes ago, Andy Perrin said:

Black flower centers appear right next to visible (only 380-400nm is enough for that). It’s not much of a test. Most lenses will pass that range, even modern ones with many elements. 
 

 

That wasn't my point. The black centers showed that here was not significant IR leakage.

[lukaszgryglicki]

I'm at the UV-B phase right now and seems like with success - I'm quite confident. UV-C seems to be out of range mostly due to filters right now, but time limits me and I need one day just to be able to play with all possible options I have.

 

[Andy Perrin]

Heh, Lukas, let's wait till we finish checking the UV-B results. Honestly, I'm leaning towards them being okay, but let's continue checking.

[lukaszgryglicki]

I will be providing more UV-B examples (holidays will finally give me some time) but the problem would be weather, I'm not sure if I can get anything useful with overcast skies...

 

[Andy Perrin]

By the look of it, it's going to be impossible to do UV-B with sunlight anyway, given what we now know about the ozone absorption in that region (from the other thread). Try with artificial light sources.

[lukaszgryglicki]

I still think we can, call it upper UV-B/lower UV-A or the lower limit that can be photographed with natural light.

Unless that peak in 280-300 is real and I can get into it as well somehow.

EDIT: one more thing - UV-B causes sunburn, so it eventually gets to the surface, so I'm quite sure I can record it, at least mixed.

 

[ulf]

I am absolutely not an expert on this. Jonathan will be much better to ask about sunburn, but I think it is a gradual thing and found this statement on the web:

 

"Exposure to longer wavelengths (360-400 nm) of UV-A radiation can lead to tanning, while exposure to shorter wavelengths (315-360 nm) can lead to sunburn. a. Provide a brief explanation as to why the shorter wavelengths of UV-A radiation are more damaging. b. What is the frequency of a photon of UV-A radiation with a wavelength of 335 nm?"

[Andy Perrin]

Well, I mean, even in the graphs we looked at there was a tiny bit in the 300-320nm range so perhaps that causes some of the burn. But the problem for you is that you have to block the 320-330 even more to use the remainder. 

[lukaszgryglicki]

I have no way of doing this, probably stack of two IV308 would do the job, but I don't have more IV 308nm filters.

 

[Andy Perrin]

Lukas, go to New Mexico in summer! At least part of the reason you have no UV-B is your time/location. :-)

[lukaszgryglicki]

Been there already, but in November :P

 

[JMC]

I'll chip in my thoughts here.

 

300nm is the effective cutoff for light getting through the atmosphere to the ground.

 

308nm imaging with sunlight can be done with a monochrome converted camera and a decent lens like the UV Nikkor. I've done it with the IV308 filter, combined with either 2mm or 4mm of Hoya U-340. Expect long exposure times and you'll need a nice sunny, summer day. If it's a day for a sunburn, then it's a good day for UVB imaging. The Edmund filters from 310nm upwards can also be used with some success, but need more blocking and a really long lens hood, as the blocking drops off axis (which I am sure I have mentioned before).

 

About skin damage, the simple definition is UVB for burning, UVA for aging (deeper penetration into the skin and more damaging to the collagen structure). As with all things it will depend on dose as well. Have a look at the erythemal action spectrum for skin damage for more information. For instance see here - https://ec.europa.eu/health/scientific_committees/opinions_layman/en/sunbeds/figtableboxes/figure-2.htm

[lukaszgryglicki]

So my images made with UV-Nikkor + IV308 + HU340 4mm thick + mono sensor with quartz coverglass have a chance to register UV-B in 300-320 area? (also including some UV-A up to maybe 330)? It was around noon, sunny day, but winter and 52 degrees north...

 

[JMC]

42 minutes ago, lukaszgryglicki said:

So my images made with UV-Nikkor + IV308 + HU340 4mm thick + mono sensor with quartz coverglass have a chance to register UV-B in 300-320 area? (also including some UV-A up to maybe 330)? It was around noon, sunny day, but winter and 52 degrees north...

 

Yes I would say the equipment has the ability to see UVB. Not so sure about the time of year though providing enough UVB to image, I've only ever tried with sunlight in summer.

[lukaszgryglicki]

Another question, one big unknown, do you have any idea if a standard camera sensor (Sony's one in 50R in this case, but I ask in general) is able to detect UV0C at 253.7nm? I mean - lens passes this, filter passes (or no filter), coverglass is quartz, there is no Bayer filter - so I assume UV-C photos hit "pixels" - the question is - will they (those pixels) detect UV-C photos or they will be absorbed on some (unknown for me) additinal glass present in pixel alone or just discarded because they are absorbed by something else before actual; detection happens? How about 222nm photos, 184nm photons?

Say 184nm is too low and cannot be detected (I assume) - what prevents "raw" pixel from detecting them? Do you have any idea?