Hey all,
So for those of you who frequent the guides section you’ll know I like to create a lot of, well, guides. But i seldom offer little insight into my research. Sure we can all Google something but getting expert advise on something specific is generally very rare or includes a price tag. As I’m not super blessed financially I usually opt for everything I can find on my own brain power. For the past few weeks I’ve been looking into UV light testing and general protection for trading cards. What I could pull up on Google was nice but it was hard to type in “how much light can a pokemon card take” only to find nothing helpful over and over again.
So I did the next best thing I could, reach out to somebody who does UV light testing for a living. To be fair, it took a while to actually find somebody who was actually willing to help me answer anything at all, but through a storm of emails I got a name: Doug Nishimura. If I may rip Doug’s bio: “…Senior Research Scientist, received his degree in chemistry from McMaster University in Canada. He is a member of the joint ISO-ANSI committee responsible for the physical properties and permanence of imaging materials. Before joining IPI as a research scientist, he did stability research at the National Archives of Canada. He is technical supervisor for most major research projects [For the Image Permanence Institute] and is well known as a teacher, lecturer, and consultant on image preservation.”
Hot damn, sounds like a guy who might know a thing or two. I fired off a very basic email asking about how he ran his light tests and what kind of lights I should use for testing (such as grow lights or a tanning bed light). I didn’t hear back for a good week and figured that was that, time to try a new avenue… then I heard back from Doug, and he sent me this:
"Hi Jason. Here we go again. There isn’t an answer to your question, but you should be able to answer your own question.
What you need to know is that you need to simulate the actual condition of use as closely as you can. A tanning light will probably be a little higher on the violet and UV end and doesn’t necessarily simulate actual daylight. A grow light for plants will be idealized for plant growth (Although I recall a graduate student studying horticulture, complain to my mother that he didn’t like grow light only for plants because they tended to cause plants to grown rather tall and spindly. I know that it would have to have a certain red component to it because the chlorophyll required for photosynthesis absorbs in the red region (which is why it appears green.) I don’t know how much UV it contains, although I had a post-graduate fellow use grow lights for light bleaching photographs and they did pretty well (other than a lot of technology failures) so there must be a reasonable amount of UV in order to photo generate hydrogen peroxide from oxygen and water.
The problem is that visible light will also cause fading, although UV tends to be a more significant threat. Ignoring exposure time, what is important is the power distribution curve of the light because it gives you power (energy per unit time per wavelength) as a function of wavelength. With light, the total energy is a function of wavelength (E=hc/l. E is energy, h is Planck’s constant, c is the speed of light, and lis wavelength) combined with what people wrongly call, “intensity” (or brightness). “Brightness” is a function of the number of photons hitting a unit area per unit time. SO the total power per unit area is the a combination of the amount of energy that each photon carries (calculated from wavelength) and the number of those photons hitting a unit area (square meter) per unit time (per second.) Energy per second is power (watts) so we get watts per square meter which is radiant flux. Visible light should be measured the same way, but because we see visible light, we use photometric units and the analog to radiant flux is luminous flux that we measure in lumens per square meter (or lux.)
What your trading card is exposed to is a function of the radiant energy hitting the trading card packaging or case and what the case glazing transmits. So, if your illuminant source doesn’t match what the cards will be exposed to, then you won’t necessarily get the correct answer that you’re looking for.
Maybe I should’ve starting by warning you that light testing is quite difficult if you want meaningful answers.
I was involved in writing an ISO light fading standard and the first one took 11 years to write and by then, it needed updating.
We use a Q Sun Xe-3-HC xenon light tester that simulates daylight through glass. We could have taken the ASTM approach (ASTM is the American Society for Testing and Materials) and they have a standard for artists colorants that simply says to put it in a sunny window. It’s cheap and certainly exposes the sample to real-life conditions, but for us, its not very meaningful because they don’t specify any methods for control. The power distribution of the light changes during the course of the day and a winter exposure is different from summer. Also cloudy, rainy or snowy weather also affects the results of the test and there’s no way to control that. There are also no specifications about the testing temperature and relative humidity and that affects the results of the test as well. Our test results won’t tell us what will happen in Florida necessarily or even Michigan, but at least we can provide the assumptions that are made about what “real life” is for the test.
There are two kinds of testing that you could do. One is comparative and, to a degree, it’s easier. Or you could do predictive, although our predictive methods aren’t very good.
I’m assuming that you’re testing methods for exhibiting or displaying the cards. Otherwise, the easy answer with no testing would be to put then into an opaque container or enclosures. So you’re looking at protective glazing of some kind (whether you use it to glaze a framed package or a display case.)
In a comparative test, you might take all of your options and expose them to the same condition and compare them. On weakness is that you only know relatively how they perform. A might be better than B which is better than C, but if A actually suffers from severe fading in four months (longer than your test period), then it doesn’t really matter much that C and B are worse. On the other hand, if C didn’t create significant fading or color shift for 400 years of real world exposure, then it’s important to know, but you may not consider that A or B is necessary either. So comparative testing has this weakness of not telling you how long things will last in each system unless you run low radiant flux tests (normal illumination and wait for as long as it will take to cause measurable fading or color shift.) It could take months or years depending on your assumptions about real life and in most cases, no one wants to wait that long.
Another option to consider is how you run the test to fit your evaluation method. There’s no right or wrong answer to this problem. You have the option of running the test for a fixed period of time and measuring the amount of change or you can run the test to a given degree of change and then compare how long it took to reach that degree of change in each sample. Each method could produce completely different results because of the way that objects deteriorate. The problem is that materials don’t change linearly with exposure to a given stress. If it did, the research and testing field would be so much easier. Instead, we might get change that appears to be more or less linear or we get things that change very rapidly initially and then they may simply slow down until they can no longer be measured or they may bottom out at some degree of change and then almost come to a stop (very slow change.) Other materials, especially if they contain a sacrificial stabilizer, show very little change initially, but when the stabilizer finally runs out, then change can be very rapid and catastrophic. Whichever path you chose, you also have to decide either how long you want to run the test (and that can change the result that you get) or what endpoint you want to use (and that can also change the result that you get.)
The other option is predictive testing, but that has its own problems. We assume that to a first approximation, that reciprocity law applies. This means that we expect to see an equal degree of deterioration for an equal level of exposure where exposure is luminous flux (“intensity”) multiplied by time. This is what camera exposure assumes, that if we close the aperture of the lens to allow only half the amount of light through, then we can compensate by doubling the exposure time. Usually it works, unless we work at extremely long or extremely short exposures.
However, it doesn’t really work in practice. Typically as we increase intensity (luminous flux), then we see a proportionate change in the time to reach a given degree of change. This works up to a point, but when we get bright enough, we find that the time is no longer proportionate, but takes longer time. The deviation from a proportionate change with intensity increases the brighter our light exposure system gets. The result is that with very high intensity testing, our predictions are longer than reality so things fade faster than we think they should and we’re misled. With chromogenic photographs, industry scientists are pretty sure that the problem is that the rate of diffusion of water and oxygen into the gelatin binder is limited. So initially, the rate of fading goes up as intensity goes up (as the number of photons per unit area per unit time goes up) but eventually, the reactions are consuming water and oxygen faster than it can be replenished my movement into the photographic binder.
Mass produced trading cards are most likely printed by offset lithography and this process doesn’t have a binder layer like photographs, but the ink colorants are bound in a binder-like material (often a drying oil.) So we expect to see the same sort of problems.
Temperature is also an issue because if we hold relative humidity constant and raise the temperature, the equilibrium water content of the card will go down and this will generally slow down the rate of fading and can give you false results. We used to use an Atlas Ci-35 weatherometer for daylight testing (we had a 50 kilolux fluorescent light fading unit as well), but the Atlas operated at 35 °C ( 95 °F) and 25% RH which caused severe drying of the samples. Our Q Sun unit has a built in refrigeration system, so we can operate at lower temperature and higher relative humidity, but we still have to account for radiant heating and that will vary from color to color. To further complicate things warmer temperatures can also speed-up reaction rate and make fading faster just by increasing the energy of molecular motion. So we’re stuck trying to figure out how to balance the slowness of dry samples versus the acceleration due to higher temperature.
It’s a difficult area of testing and research to deal with and I can tell you that after 35 years of research, I don’t have the answers yet.
Oh. You’ll have to figure out what you want to test for as well. I expect that the cards are probably not very pure cellulose and therefore contain a lot of lignin which can cause significant staining. Older cards on acid (sulfite) processed pulp cardboards may also suffer from physical deterioration and become more brittle. However, you may be most concerned with fading and real world experience with posters in bank and restaurant windows show us that more than a few months of window exposure can cause pretty significant fading.
Good luck.
-Doug"
For those of you who want a TL;DR version: Light testing is so frightfully difficult even scientists who have done this all their lives still don’t do it correctly. There are a ton of variables to account for but in the end your probably better off just using the sun to test any kind of fading.
So there you have it! I’ll be attempting to do some light testing myself in the future and we’ll see how it goes, thanks to Doug’s email I now have two tests that I would like to run and I’ll be posting guides relating to that in the coming months, possibly sooner depending on what kind of results I get.
I’ll be at worlds in a few weeks, if I meet any of you there I saw we raise a glass to Doug as he’s given us a direction to go in terms of learning how to protect our collections.
Thanks for reading!
