Thorium glass?

[reddesert]

I think that's a reasonable guess. I'd add one more thought to it: photon energy. This correlates with wavelength and blue photos are thus less energetic than ones in the UV bands. There may be a threshold energy required to bump the stray electrons back into place and allow the matrix to relax. This likely depends on the material and perhaps on the type of F-center. I fed the question into AI (sketchy, of course) and it comes up with a range of 2.5-3.5eV, which encompasses the longer UVA wavelengths and the blue spectrum. It also alleges that UV is more efficient; this seems to make sense when taken at face value given the higher energy involved, which would make it more likely that an exposure event has the desired effect. I think it's a reasonable assumption that UV will work better, but that blue might get the job done provided sufficient exposure is given.

Yes, energy has to matter because if a photon isn't energetic enough it won't kick whatever is causing the defect over the energy barrier to put it back into place. I remain a little skeptical that the effect is exactly analogous to F-centers (electrons sitting in an atom's spot) in crystals, because glasses aren't crystals but amorphous solids, and so have a much less regular structure even at the atomic scale.

This is possibly relevant because I suspect the irregular nature of the solid makes the defect energy a less well defined quantity. In other words, for some types of solid-state phenomena, there's a fairly strict excitation energy and a photon below that energy just isn't going to do anything. (A similar-but-different example is the band gap in semiconductors. Which is tangentially why you can't use a silicon CCD to detect near-infrared photons, they aren't energetic enough.) For crystals, if you displace a single atom from one lattice location or another, the energy should be about the same since the periodic structure is the same. But for glasses, the energy might be different between two locations, since even the surrounding short range structure is different. However, higher energy photons will definitely be more likely to have an effect: UV beats blue, and who knows if redder light is energetic enough.

Anyway, that's hand-waving. A little bit of web searching suggests that the glass transition is not as easily modeled as crystalline structure. I got more specific and found (by googling eg "activation energy of defects in silicate glass"), this article: https://www.sciencedirect.com/science/article/abs/pii/S0168583X98008842
"Activation energy spectra for annealing of ion irradiation defects in silica glasses" which seems topical. I can't read the whole thing (possibly when I get back to the office). They cite defect energies of 0.26-1.85 eV in alkali-borosilicate glasses, which are at least similar to optical glasses. I believe the radiation intensity used to create the defects was much much higher than the kind of self-irradiation familiar in thorium glasses.

 

[koraks]

If we had a thumbs-up feature for posts, I'd have given yours one @reddesert!

 

[Petrochemist]

Yes, energy has to matter because if a photon isn't energetic enough it won't kick whatever is causing the defect over the energy barrier to put it back into place. I remain a little skeptical that the effect is exactly analogous to F-centers (electrons sitting in an atom's spot) in crystals, because glasses aren't crystals but amorphous solids, and so have a much less regular structure even at the atomic scale.

This is possibly relevant because I suspect the irregular nature of the solid makes the defect energy a less well defined quantity. In other words, for some types of solid-state phenomena, there's a fairly strict excitation energy and a photon below that energy just isn't going to do anything. (A similar-but-different example is the band gap in semiconductors. Which is tangentially why you can't use a silicon CCD to detect near-infrared photons, they aren't energetic enough.) For crystals, if you displace a single atom from one lattice location or another, the energy should be about the same since the periodic structure is the same. But for glasses, the energy might be different between two locations, since even the surrounding short range structure is different. However, higher energy photons will definitely be more likely to have an effect: UV beats blue, and who knows if redder light is energetic enough.

Anyway, that's hand-waving. A little bit of web searching suggests that the glass transition is not as easily modeled as crystalline structure. I got more specific and found (by googling eg "activation energy of defects in silicate glass"), this article: https://www.sciencedirect.com/science/article/abs/pii/S0168583X98008842
"Activation energy spectra for annealing of ion irradiation defects in silica glasses" which seems topical. I can't read the whole thing (possibly when I get back to the office). They cite defect energies of 0.26-1.85 eV in alkali-borosilicate glasses, which are at least similar to optical glasses. I believe the radiation intensity used to create the defects was much much higher than the kind of self-irradiation familiar in thorium glasses.

Activation energies apply to all chemical reactions not just crystalline ones.

The radiation that creates the defect (yellowing) will come from an alpha particle emitted as the Thorium decays. These are highly energetic & react quite readily (so don't travel far typically only a couple of inches in air). The yellow species created is probably only semi stable - higher in energy than the original clear species & only needing a much less energetic stimulation to get back over the energy ridge. The absorption of blue light in stimulating this reverse reaction would actually explain the yellow colours too.

FWIW you can used silicon based CCDs for NIR upto ~1100nm It's above that wavelength that silicon effectively becomes transparent to IR.
I've taken loads of NIR images with a silicon based CCD, and also silicon based CMOS sensors. Silicon is quite useless for thermal IR which is what many people think off if you mention infra red.
I've been careful to say silicon based as it may be that changes to the regular doping are required to get suitable reactions to red light (essential in colour cameras). Digital cameras routinely fit a hot mirror before the sensor to limit UV & NIR wavelengths that the sensor would otherwise see.

 

[AZD]

Anyhooo… another dandy thorium lens followed me home yesterday - a neat Mamiya Sekor 55mm f/1.4 M42 with Spotmatic still attached as a rear lens cap. Couldn’t resist for $25. Not too much yellowing on this one so I’ll leave it alone for now.

Looking forward to running it through the paces as there’s not a whole lot of examples online.

Good to know that if the yellowing ever bothersome I have a convenient bandwidth of frequencies to select from when I get the F out of them F-centers.

 

[blee1996]

My newly acquired Aero Ektar 6" is a bit yellow, so I will look for some proper UV lights. The Ikea Jansjo LED lamp did not make much differences after 1 week with my previous Takumar.

 

[AZD]

My newly acquired Aero Ektar 6" is a bit yellow, so I will look for some proper UV lights. The Ikea Jansjo LED lamp did not make much differences after 1 week with my previous Takumar.

At the hardware store, big box store, or whatever, look for an “EcoSmart 40-Watt Equivalent A19 Blacklight Ultraviolet Glow in the Dark LED Light Bulb”. It has worked faster than anything else I’ve tried.

 

[blee1996]

At the hardware store, big box store, or whatever, look for an “EcoSmart 40-Watt Equivalent A19 Blacklight Ultraviolet Glow in the Dark LED Light Bulb”. It has worked faster than anything else I’ve tried.

Thank you for the tip. I will look for those at Home Depot.

 

[blee1996]

At the hardware store, big box store, or whatever, look for an “EcoSmart 40-Watt Equivalent A19 Blacklight Ultraviolet Glow in the Dark LED Light Bulb”. It has worked faster than anything else I’ve tried.

Thanks for the suggestion: the Ecosmart blacklight UV bulb works well! After 48 hours, the yellowing is mostly gone. I installed the bulb in a desktop lamp housing, put the rear group of the lens in a Danish cookie tin jar (to reflect more light from bottom up), and kept the bulb 1-2 inches above the lens. The bulb/lamp runs pretty cool to the touch, but I still turn it off for 15 minutes every 24 hours. And unlike specialized 365nm UV penlight, this light is not harmful to your eyes unless you stare at it for extended period of time.

 

[titrisol]

I have looked up technical references on radiation damage in glass a couple of times - one is linked in this thread https://www.photrio.com/forum/threads/yellowing-of-pentax-1970s-lens.217074/post-2948462
but unfortunately you need an institutional login like a university library to see the article - and the references I've found haven't given much detail on the mechanism of how the absorption of light undoes the damage to the glass structure. It seems likely that the absorption excites out-of-place atoms and they relax back into place, but the specifics aren't explained.

My guess is that since the color of the dislocations absorbs blue-ish light as well as UV, blue-ish light may result in enough absorptions to eventually repair the damage, but I have no empirical tests. I do have a yellowed Takumar lens and may someday try to expose it to different kinds of light each for some time and see if the rate of increase of optical transmission can be measured, but it is likely that with home equipment I would not be able to make measurements precise enough to be definitive.

I have the article, let me know how to send it to you
And yes, a SuperTakumar 50/1.4 that I havent used since before 2012 has about 1/2 stop of yellowing
The S-M-C Takumar cleaned about 5 years ago is not noticeable

 

[AZD]

Thanks for the suggestion: the Ecosmart blacklight UV bulb works well! After 48 hours, the yellowing is mostly gone. I installed the bulb in a desktop lamp housing, put the rear group of the lens in a Danish cookie tin jar (to reflect more light from bottom up), and kept the bulb 1-2 inches above the lens. The bulb/lamp runs pretty cool to the touch, but I still turn it off for 15 minutes every 24 hours. And unlike specialized 365nm UV penlight, this light is not harmful to your eyes unless you stare at it for extended period of time.

Glad it worked. I had a similar setup, but using the lamp housing of an old Federal diffusion enlarger with the lens placed inside. Seems like I also saw the most change in the first 2 days.

 

[Petrochemist]

Glad it worked. I had a similar setup, but using the lamp housing of an old Federal diffusion enlarger with the lens placed inside. Seems like I also saw the most change in the first 2 days.

That makes sense. The UV that reverses the yellowing reaction is absorbed by glass, so yellow near the surface will react fairly quickly, then the stuff further down will be slower as it's getting less UV dosage.