Impact of Light on the Drying of Oil Paints: Initial Testing

Oil paints dry through oxidization, so access to air seems clearly important, but what about light? How much does exposure to different light levels impact the drying process? Surprisingly, for a question that seems so basic and fundamental, there is not a lot of information or test data that one can find. While a few scattered studies do turn up in the literature, they are focused almost solely on the oil alone and are not based on a range of typical light levels and a variety of common colors. What follows describes some of the preliminary testing we have done in this area, and although there were no clear and definitive conclusions, we did end up with some tantalizing curiosities and patterns that can help inform best practices and provide areas for further investigation.

Test Layout

The testing included 10 Williamsburg Oil Colors chosen from three different drying categories and including both reactive and inert pigments:

  • Fast: Raw Umber, Phthalo Green
  • Medium: Cadmium Red Medium, Cobalt Blue, Ultramarine Blue, Titanium White, Titanium-Zinc White, Flake White
  • Slow: Quinacridone Magenta, Lamp Black

These were applied as 6 mil drawn downs, about the thickness of two sheets of common copy paper, onto uncoated polyester film.

Samples were then placed in one of 8 different lighting situations. All lighting involved standard fluorescents, although the different types (daylight, cool, full=spectrum, etc.) were not controlled for. Lux levels, or the luminosity and brightness of the light, were measured using an Extech LED Light Meter.

Lighting situations and the corresponding lux levels:

  • Brightly lit table directly under light fixture (1050 lux)
  • Brightly lit wall adjacent to table (840 lux)
  • Office wall facing window (580 lux)
  • Drying shelf in Lab (370 lux)
  • Moderately lit wall (350 lux)
  • Open shadow on brightly lit wall (185 lux)
  • Open shadow on moderately lit wall (60 lux)
  • Completely dark room (0 lux)

Weekly totals were then estimated based on the number of hours those particular lights would be on during a normal week.

Test areas. A) Bright Table, Bright Wall with open shadow above lights B) Office with Window C) Moderate Wall with open shadow in upper left D) Lab with drying racks

Temperature and humidity were not controlled for, so any differences at the various locations were not taken into account.

A separate parallel study involved 6 mil drawdowns of Titanium-Zinc, with one sample placed in the dark room and additional samples transferred there after being exposed to either direct daylight or hung on the brightly lit wall for various intervals. This was to see if initiating the drying of a sample in the light impacted the length of time it took to finish drying in the dark.

Test Results

The test results, as you will see, contained more than a few surprises. Some colors were unaffected by the amount of light, drying at the same rate regardless of location, others showed a very moderate increase, while for a few the differences were dramatic.

First up are the five colors that were, for the most part, invariable: Cobalt Blue, Raw Umber, Phthalo Green, Ultramarine Blue, and Lamp Black:

This is an unusual group by any measure. While the pigments in three of the colors (Cobalt Blue, Raw Umber, Phthalo Green) contain elements that can act as driers, such as cobalt, manganese and copper, making it is easy to think there might be an internal mechanism that allows them to kickstart the drying process on their own, this is not true of Ultramarine Blue or Lamp Black. Indeed Lamp Black is a notoriously slow drier and chemically inert, so one would assume it would be greatly impacted by light exposure, but it was amazingly consistent regardless of the amount of light.

The second group is made up of just two colors, Flake White and Quinacridone Magenta. Under most circumstances, both these colors had fairly consistent dry-times, but showed a clear increase when placed in the dark or even in very subdued light, as in the case of Quinacridone Magenta. The variations in some of the data is most likely due to differences in sample preparation and where a particular sample was placed; although the sudden drop in drytime for the  Quinacridone Magenta located in the Lab remains an unexplained anomaly.

Finally, the last three colors (Titanium White, Titanium-Zinc White, and Cadmium Red Medium) exhibited a strong correlation between light levels and dry times. As the light levels grew lower, the drying times steadily increased in a gradual and logical manner, ending with a very sharp spike in dry time when placed in the dark:

While both Titanium White and Titanium-Zinc dried 3 times slower in dark storage than the open shadow of a moderately lit wall, jumping from 7 to 22 and 28 to 90 days respectively, Cadmium Red Medium displayed an extraordinary increase in dry time, going from 16 to 135 days, or about 4 1/2 months.

Implications and Next Steps

We undertook this testing with the common assumption that light levels would impact dry times in some regular, predictable way that would be easy to draw conclusions from. However, the results have ranged over a wide area, and given the small sample size, there is no clear, singular factor that explains why some colors are in one grouping or another. Also, what was most surprising was not the dramatic jumps in dry time when colors were placed in the dark, but those cases where light levels or even the absence of light appeared to make no difference whatsoever. While pigments appear to play the central role in these differences, much more testing will be needed to understand how. And naturally, we are left with other questions as well: How do mixtures of different colors behave? Does layering one color over another make a difference? Are all the results similar under different lights – fluorescent, halogen, incandescent and LED? Is the presence or absence of UV important? These are some of the areas we hope to explore in the future, in addition to expanding the number of colors and types of pigments being tested.

Best Practices

While it is hard to construct best practices based on limited testing, it still seems reasonable to assume a few things that can help your studio practice:

  • Paintings placed in the dark or in subdued lighting might dry significantly slower. This includes when paintings are placed in drying racks or turned against the wall.
  • Conversely, paintings exposed to ample amounts of light should dry more quickly, although which types of light are the most effective still needs to be tested.

However, because pigments appear to have a large impact on the results, the above points are at best broad, general guidelines. Individual paintings are invariably complex, with multiple pigments being blended together and mixed with various mediums. Exactly how all of that impacts the drying times is an area that is just now being explored for the first time.


13 Responses to Impact of Light on the Drying of Oil Paints: Initial Testing

  1. Charles April 24, 2018 at 9:51 am #

    Until you you can separate the effect of heat (temperature) on oxidation of oil paints, you will not really know the effect of light on oxidation. Your test results indicate the combined effect of heat and light.

    Do you have any test results on the effect of temperature on oil paint drying (oxidation)?

    • Sarah Sands April 25, 2018 at 10:24 am #

      Hi Charles –

      Thanks for the comment. I agree temperature is a variable that needs to be accounted for to get a complete picture, and since heat accelerates all chemical processes, there is every reason to believe that it will speed the drying times across the board. But still important to understand to what degree and the amount of impact versus light exposure. As you can imagine, it is a little tricky to control for temperature in a test, while keeping other variables constant, but we are working on some test designs that will allow us to try to do some of this in future rounds. For now, as I stated in the article, the most surprising thing is not that light had an impact, but that a full half of the colors exhibited no or little change in drying time even in complete darkness, with temp at typical ambient conditions as might be found in most studios and homes. That more than anything is what needs to be explained and understood.

      But more to come… will take a few rounds to really suss this out.

  2. Michael Rothman April 24, 2018 at 10:40 am #

    Dear Ms. Sands,

    I applaud this line of inquiry. However, one issue in the experiment design stuck out to me: the method or interpretation of what constitutes a dry film, when it is present. You didn’t mention how the swatches were sampled or tested for evidence of dryness. Say, as opposed to “skinning”. Were the samples touched with some standardized probe or just the finger of a technician?

    I am looking forward to hearing about future discoveries your and your team at Golden/Williamsburg might make.

    Best wishes,
    Mike Rothman
    Illustrator and

    • Sarah Sands April 25, 2018 at 10:32 am #

      Hi Michael –

      Thanks for the comment and great question as that is important and likely should have been noted. My oversight on that. Our dry time tests are pretty aggressive in that we take a toothpick and draw it across the surface with constant, reasonable pressure and consider it dry when it can go across the entire width without going through the paint or causing it to rip. This gets around the fact that paints will dry along the edge before the center, and gets one past the more ambiguous ‘touch dry’

  3. Bev Laird April 24, 2018 at 10:44 am #

    I’m surprised there was no inclusion of sunlight in the testing format. I know of several oil painters who have constructed UV light boxes with special UV-emitting lights to speed the drying of their oil paints.

    • Sarah Sands April 25, 2018 at 10:48 am #

      Hi Bev –

      We did include some trials with sunlight, and got some interesting results, but wanted to expand that section more before writing anything up. So look for that to be discussed in a future article. For now we just wanted to share these initial results, which were meant to establish a baseline using a range of light levels that would be typical indoors. But by no means was this meant as definitive – just a first foray into a topic that has surprisingly remained unexplored in a controlled way.

      Also, as I point out, what is surprising is not so much the fact that some paints dried faster with more exposure to light, but that a full half of the samples dried just as well in complete darkness with no light or UV exposure. That runs counter to most expectations and there is no easy common connection between the pigments that would explain this behavior. So a lot to look at. But definitely as part of future rounds, we will be including sunlight, along with various types of other light sources (LED, halogen, incandescent, and blacklight). So stay tuned!

  4. Joseph Sundwall April 24, 2018 at 12:04 pm #

    Interesting. But no direct sun data? What else I’ve always wondered over is the effect of heat and forced air… perhaps you’ve already done those studies?

  5. Brent Ciccone April 24, 2018 at 12:24 pm #

    The effect of direct sunlight would be interesting to see. Also, the question of different drying times for different pigments, how does that affect the paint film strength, and does it lead to eventual cracking of the paint surface when parts of the painting are drying faster than others? Should our goal be an even drying time of all different colours rather than a faster drying time?

    • Sarah Sands April 25, 2018 at 11:01 am #

      Hi Brent –

      Thanks for the comment and great questions! Empirically we know there are far fewer problems with different drying times in alla prima painting, where everything is applied in one sitting and dries all of a piece, versus the complications when layering paints on top of each other. However, your questions definitely point to areas that need more testing. Unfortunately, as you likely know, it can take years for cracking and any meaningful differences in film strength to show up, which makes test design a bit challenging. But will definitely add these questions to our list of ones to explore.

      As to sunlight, as I mentioned in another comment, we did do some very limited testing in this area, and got some interesting results, but wanted to expand the scope before writing anything up. So look for that to be discussed in a future article.

  6. ralf r April 25, 2018 at 4:13 am #

    indirect bright indirect sunlight (either art studio indirect light or direct but reflected light) on the slower driers would be very interesting in the next study

    • Sarah Sands April 25, 2018 at 11:06 am #

      Hi Ralf –

      More than any other request for future testing, adding sunlight as a variable is on the top of the list. So you are not alone! Look for that area to be explored in more depth in future rounds. In the meantime, keep in mind that at least one site in these tests had a window not far from the wall, so at least some of the data involved samples with indirect sunlight as a component, and those dry times were in line with the others. But more needs to be done in this area to really see the impact of direct and indirect sunlight exposure.

  7. Leonardo Puig April 29, 2018 at 7:07 am #

    Thank you very much for the experiment and for sharing it. In my personal experience I know that the paint on the canvas and the palette in summer dries much faster than winter. I already know that in summer there is more light, but I am referring to a room with artificial light where natural light is only a complement. Here the ambient temperature and the low humidity surely help. On the other hand when I want to preserve a palette with colors that I have pre-mixed because I need to suspend the painting session or protect some color to make corrections in the next few days in the painting, I cover my palette with PVC film and store it in the fridge. This works by combining less air through the pvc cover, darkness, low humidity and cold. Greeting.

    • Sarah Sands April 30, 2018 at 12:51 pm #

      Hi Leonardo –

      Thanks for the comments. Definitely temperature should have an impact since heat will generally speed up chemical processes. And of course cold will slow things down. As a factor humidity is less clear since oils are not drying through evaporation, so the level of saturation of water vapor in the surrounding air should not have an impact. But of course in the real world it is hard to separate humidity from other factors like temperature, so it is usually difficult to separate out all the factors. Finally, while we know that many people will store their paints and palettes in a frig or even freezer, this is not something we can recommend or endorse as it places paint alongside food items. If wanting to do this regularly it would be best to have a dedicated small fridge for purely this use.

Leave a Reply


Made by Golden Artist Colors, Inc.